Table I.
Summary of Results
1 Rod temp., Rod surface area, cm. * Carbon evap. rate, g./cm.* secb Totalcarbonevaporated, g. Concn. identified products,
2
1893 7.5
2122 6.6
3.3
x
lO-l3
8.4X
8.9
x
10-9
2 . 0 X 10-6
3
4
5
6
7
2225 6.6 2 . 3 X 10-10
2253 7.5 2 . 3 X 10-9
2320 6.6 8 . 8 X 10-9
2353 7.5 1 . 6 X lo-*
2390 6.9 3 . 5 X 10P
4.3 X d
8 . 7 X IO-’ d
d
d
1.8 X 2 . 1 x 10-5
6 . 2 X 10-5 4 . 9 x 10-5
2.1 X 1 . 1 x 10-4
d
d
-2
8.6 X
4.2 X
A4c
Concn. toluene Concn. cycloheptatriene
X
d
(I
The carbon evaporation rate data used were from the work of M. Hock, P. Blackburn, D. Dingledy, a Measured by optical pyrometry. and H. Johnston, J . Phys. Cliem,, 59, 97 (1955). c Product yields were determined by comparison of peak heights from standards; peak Product heights for toluene and cycloheptatriene are linear in concentration over the concentration range 3 X 10-l to 3 X 10-5 M. M for toluene and 1.8 X 10-5 ik! for cycloheptatriene for 2-fil. injections. yield was less than the limits of detection, which are 8.8 X e This yield was just barely above the limit of detection.
of bombarded benzene was collected by vaporization and condensation in a liquid nitrogen cooled cold finger, which was sealed under vacuum. Chromatographic product analyses were performed using silicone oil columns a n d flame ionization detection. Products were identified by comparison of their relative retention times (relative to the solvent benzene peak) with those of knowns in benzene solution. The experimental results are summarized in Table I. As is shown in the table, despite a constantly increasing rate of carbon deposition the yields of toluene and cycloheptatriene increased for only the first four runs before falling below their detection limits during the next three runs. Further, whenever the total carbon deposited per irradiation was greater than 3 X 10W g.4 (i.e., runs 4 thru 7), a dark brown material was observed to form on the walls of the bombardment flask during the course of the experiment. Qualitatively, the yield of this material, which was nonvolatile at 1 p even when heated to ca. 50”, increased rapidly with increasing carbon deposition per irradiation. Thus, as the yield of this material increased, the yields of toluene a n d cycloheptatriene fell below their limits of detection. That the toluene and cycloheptatriene were not produced by photolysis or pyrolysis of the benzene under the conditions of our experiments is demonstrated by runs 1 a n d 2 at 1893°K. and 2122°K. In these runs carbon evaporation rates were t o o low to lead to detectable product levels. However, the operating temperatures of run 1 and particularly of run 2 were not sufficiently different from our normal operating temperatures to effect significantly the yields of products formed by photolysis or pyrolysis. Thus, the failure to detect toluene or cycloheptatriene in either run eliminates the possibility that they were formed by photolysis or pyrolysis. Skell has reported previously that Ca from carbon vapor reacts with isobutylene to form 1,1,1’, 1 ’-tetramethylbi~ethanoallene.~The present work provides the first example of reaction of Ci from carbon vapor with a n organic substrate.5a Nuclear transformations have been used extensively (4) It is interesting to note that this amount of carbon contains approximately the same number of carbon fragments as there are surface benzene molecules. ( 5 ) P.S. Skell and L. D. Wescott, J . A m . Chem. Soc., 85, 1023 (1963). (5a) NOTEADDEDI N PROOF. Skell and Engel [ibid., 87, 1135 (1965)l have just reported the reaction of C I from a carbon arc with cis- and trans-2-butenes.
to investigate the reactions of hot and moderated carbon atoms with organic substrates.6 H o t carbon atoms have been found to react with benzene to form radioactively labeled toluene.’ Accelerated carbon ions, C+, with kinetic energies as low as 45 e.v., have been shown by Mullen to react with benzene to form both toluene and cycloheptatriene.5 Our results provide the first demonstration of a reaction between an orgaqic substrate and carbon atoms produced with an initial kinetic energy below 0.3 e.v. Methylene produced by the photolysis of diazomethane reacts with benzene to give toluene and cycloh e ~ t a t r i e n e . ~It is attractive to visualize carbene-like insertion and addition steps in the reactions of carbon atoms with benzene. Indeed, Wolfgang has postulated6b and demonstrated6b insertion by carbon atoms into C-H a n d ethylenic double bonds. However, we cannot say where, along the reaction pathways leading to toluene and cycloheptatriene, d o the necessary hydrogen abstractions take place. Also, we cannot yet offer a proven explanation for the elimination of the toluene and cycloheptatriene yields and for the formation of the dark brown nonvolatile material when too much carbon (>400 pg.) is deposited in the bombardments. (6) (a) See review by A. P. Wolf, Advan. Phj,s. Org. Chem., 2, 201 (1964); (b) M. Marshall, C. MacI
0: 3 V
z
100
0
+
5 50
a c
0 K
200
a
4 4
0
I
500
)50 100
I
I WO
240
250
260
270
280
290
300
WAVELENCITH ( m p )
Figure 1. Ultraviolet spectrum in HnO and 0 . R . D curves for Lphenylalanine. Curve A : 1.39 x lo-* in 1 N aqueous HCI (I-cm. cell); curve B: 1.26 x M i n HIO (IO-cm. cell). Vertical lines indicate random experimental error within each set of data. In the case of curve B, because of the small measured rotations due to high dilution, there is also a nonrandom base-line uncertainty corresponding to a possible shift of the entire curve estimated at less than &SO0.
weak 260 m p absorption band in the biologically important cr-amino acid L-phenylalanine (I). The point of conflict is of some significance and warrants settling, for an accurate description of the optical rotatory dispersion (O.R.D.) and/or circular dichroism (C.D.) of the a-amino acids is a prime prerequisite for the structural interpretation of protein and polypeptide optical activity data. The Schellmans, using Drude-type analyses, predicted optically active aromatic absorption bands for aromatic amino acids: for L-tyrosine (11) at 272 m p (0.66 M HCI) and for L-phenylalanine (I) at 253 m p (0.22 M HCI). The pertinent Cotton effect for I1 is now well documented experimentally, 3-fi but workers in the United States have heretofore failed to affirm its existence for Hooker and Tanfords in particular have emphasized this contrast in the behavior of 1 as compared to I1 and some of its derivatives. It is the purpose of the present note to point out that fundamental considerations governing natural optical activity require that this apparent difference in behavior be not one in kind, but only in degree, and that the ap( 3 ) 41. Billardon, Compf.rend., 251, 535, 1759 (1960). (4) E. Iizuka and J . T. Yang, Biochemistr.r, 3, 1519 (1964). ( 5 ) T. M . Hooker and C. Tanford, J . A m . Chem. So?., 86,4989 (1964). (6). G . D. Fasman, E. Bodenheimer, and C. Lindbloa, Biochemistr.r, 3, 1665 (1964): S . Beychok and G. D . Fasman, ibid., 3, 1675 (1964). (7) Billardon published data showing anomalous dispersion in the 260 m p region for 1. I n particular, he reported a Cotton effect at the acid p H 1 . 1 k i t h a peak-trough molar rotation difference of -200". Such rcsults difTer markedlb from our o w n findings and those of other workcrs cited in the present note.
1814
Journal oftlie American Chemical Societj
1
parent behavioral difference really amounts to a question of orders of magnitude that can be qualitatively understood in terms of the local symmetry of the "inherently symmetric, but asymmetrically (more generally, dissymmetrically) perturbed"* phenyl chromophore. Toward this end we also present unequivocal experimental evidence for the existence of a 260 mp Cotton effect for I. O.R.D. curves (Cary Model 6 0 ) and an ultraviolet spectrum (Cary Model 15) for chromatographically pure 1 from Mann Research Laboratories (Lot 4515) are shown in Figure 1. The O.R.D. curves are plotted on an uncommonly expanded ordinate scale; the visible and near-ultraviolet O.R.D. is so dominated by the transitions below 230 m p that the extremely weak Cotton effect would be barely discernible on a scale sufficient to display adequately the long wave length singlet aromatic Cotton effect for 11. In strongly acid solution (curve A) the Cotton effect is superimposed on a very steep background rotation that almost completely obscures the O.R.D. extrema so prominent in curve B. This explains, in part, why Iizuka and Yang4 could not decide unequivocally as to the presence of anomalous dispersion for 1 in the 260 m p region. Best resolution was obtained for dilute solutions M ) in IO-cm. cells with maximum absorbance -2.5 at the peak absorption near 257 m p (pH 5.9). Repeated runs of the undisturbed sample with the pen response time set at 30 sec. were made in order to average the noise. T o test for the possibility of artifacts d u e to stray light, we measured the O.R.D. of a solution containing L-leucine (111) and DL-phenylalanine (IV) that in the 250-300 m p region had the same background O.R.D. and the same absorbance as the samples used to obtain curve B of the figure; we could detect no Cotton effect in the 260 m p region for the 111-IV solution. To understand the extreme weakness of the 260 m p Cotton effect in I , consider a benzene ring substituted such that the effective local symmetry is C2". If one constructs the 7r-molecular orbitals of the system solely from linear combinations of 2p, atomic orbitals of carbon, then all x-x* transitions from the ground state have their nonvanishing electric dipole transition moments pe directed in the plane of the ring, and their nonvanishing magnetic dipole transition moments pm directed perpendicular to that plane.g Hence n o amount of mixing of the x-r* transitions among themselves can lead to a nonvanishing scalar product pe.pm, and hence to optically active x-x* transitions. In other words, the dissymmetric molecular environment provided for the aromatic chromophore by the amino acid substituent in I and I1 is ineffective for the generation of a x-x* Cotton effect if x-states of the sort just described are the only states involved. Further, if one can ignore the possibility of charge transfer between or the mixing of transitions between the aromatic and amino acid groups, then the only way the requisite in-plane p m can be achieved is by mixing the T-T* transitions with some perpendicular transi(8) A. Moscowitz, Tefrahedron, 13, 48 (1961); A . Moscowitz, I
